Examining the relationship between multi-agent immunosuppressive therapy for immune-related adverse events (irAE) and infectious complications.

2658 Background: Treatment of severe irAEs with multiple immunosuppressive therapies (ISTs) decreases the morbidity and mortality of these conditions. Nevertheless, the rates of and risk factors for infectious complications in this population are not known. Methods: We conducted a retrospective study of patients (pts) who received an immune checkpoint inhibitor (ICI) and experienced ≥1 irAE requiring treatment with corticosteroids along with at least two lines of steroid-sparing ISTs administered either concurrently or within 90 days of each other. We annotated all infections from ICI start until 90 days after IST. Opportunistic infections (OIs) were defined as herpesvirus (CMV, EBV, VZV, and HSV) and invasive fungal infections. Infection density was reported as number of infections per 1000 patient-days and graded as mild (not requiring treatment), moderate (requiring oral treatment), severe (requiring hospitalization or parenteral treatment), life threatening, or fatal. Risk factors were identified using univariable and multivariable Cox regression analysis adjusting for age at ICI initiation, sex, ICI regimen, ISTs, and steroid dose. Results: 175 pts (52% male, mean age: 66) with 238 irAEs and 417 ISTs were analyzed with a median follow-up of 367 days. The associated ICI regimens included αPD-1 (n = 81, 46%) and αPD-1/αCTLA-4 (n = 67, 38%). The most common irAEs were colitis (n = 67, 38%), hepatitis (n = 42, 24%), and myocarditis (n = 26, 15%). The most frequently used ISTs were mycophenolate mofetil (n = 92, 53%), infliximab (n = 77, 44%), and vedolizumab (n = 54, 31%). 103 pts (59%) developed 223 infections (median 2/pt, range: 1-8). 93 pts had 187 non-OIs. Of the 87 pts (50%) who had OI testing, 29 (33%) had 36 OIs, most commonly EBV DNAemia (n = 14, 16%) and CMV reactivation (n = 12, 14%). 6 pts had >1 OI, including 1 pt with CMV, EBV, and HSV. OI density significantly increased after starting ISTs for irAEs, but non-OI density was unchanged (Table 1). αCD20 use was associated with increased non-OI risk (HR: 10.58, 95% CI: 3.64-30.72, p < 0.001), while there was a trend towards increased non-OI risk with a max prednisone dose >100mg (HR: 1.74, 95% CI: 0.96-3.13, p = 0.07). In contrast, a max prednisone dose >100mg was associated with increased OI risk (HR: 2.89, 95% CI: 1.21-6.90, p = 0.017). 58 pts (33%) had severe or life-threatening infections, of whom 16 (9%) had OIs. 8 pts (5%) had fatal infections in this population. Conclusions: Use of multiple ISTs for severe irAEs is associated with increased rates of opportunistic infections as well as a 5% infection-related mortality rate. Patients requiring multiple lines of ISTs must be closely monitored for infectious complications, and prophylaxis should be considered when appropriate. Mean OI and non-OI density per 1000 patient-days while on ICI alone vs ISTs. ICI ISTs p-value OI 0.03 1.70 0.002 Non-OI 10.70 8.98 ns